This ebook is available only for subscribers.
In this exclusive subscirber-only ebook you’ll learn how the emissions from individual AI text, image, and video queries seem small—until you add up what the industry isn’t tracking and consider where it’s heading next.
by James O’Donnell and Casey Crownhart May 20, 2025
Table of contents
Related stories:
Stay ahead with more perspectives on cutting-edge power, infrastructure, energy, bitcoin and AI solutions. Explore these articles to uncover strategies and insights shaping the future of industries.
Those migration challenges are on top of the alleged legal intimidation and cease-and-desist letters sent earlier this year to customers with perpetual licenses who had not signed up for subscriptions. Signaling its determination to challenge Broadcom’s regime, in July ECCO also filed an action with the General Court of the
“Things are pretty straightforward for the managed cloud option — there are two obvious choices. Companies can go with IBM and Red Hat, or choose VMware,” he said, noting that while the company has been a big VMware reseller, “IBM had to ask themselves, ‘How are we going to compete
The startup designed its architecture to meet enterprise-grade security requirements. The platform is SOC 2–compliant, processes all data in memory, and doesn’t store customer logs or telemetry. That approach, the company says, allows adoption even in highly regulated industries where data control is critical. The company says the transparency is
One such AI application, developed by MITRE manages wireless spectrum allocation within a cell site in real time so users can experience seamless connectivity and operators can see improvements in spectral efficiency, Nvidia stated. Unlike traditional approaches that shut down entire network bands when interfering systems are detected; this application
Oil increased, snapping a three-day losing streak, as investors assessed a large drop in US inventories and the impact of Western sanctions against leading Russian crude producers. West Texas Intermediate rose by 0.6% to settle below $61 a barrel, while Brent closed near $65. US crude stockpiles declined 6.9 million barrels last week, the most since early September, along with draws in gasoline and distillates, according to an Energy Information Administration report Wednesday. The report pushed up oil prices to intraday highs as traders reconciled a tightening supply outlook in the West with mounting threats to Russian flows. Gasoline futures surged 2.5% on signs of unseasonably high demand. The data compounded bullish sentiment after US President Donald Trump said he would follow through and enforce harsh new sanctions against Moscow to pressure Vladimir Putin into negotiations to end the war in Ukraine, according to Matthew Whitaker, the US ambassador to NATO. Indian state-owned refiners are considering whether they can continue to take some discounted Russian oil after the measures were imposed, though some processors will pause purchases for now. Indian Oil Corp. said on Tuesday it is “absolutely not going to discontinue” purchases of Russian crude as long as it complies with international sanctions. “The market is now trying to assess the longer-term impact of the additional sanctions, which will be determined by the quantity of actual barrels removed from supply,” Standard Chartered analysts including Emily Ashford said in a note. Oil is on track to notch a third monthly decline, with prices dragged lower by expectations of a global surplus as OPEC+ raises production. Key alliance nations are set to hold discussions this weekend and may sign off on another supply increase. Adding a ceiling to prices, Federal Reserve Chair Jerome Powell said that a rate cut in December
Out of the top 25 highest energy-consuming countries, the United States now ranks 11th in energy efficiency policy and performance, according to the American Council for an Energy-Efficient Economy’s 2025 assessment. ACEEE previously ranked the U.S. 10th in its 2018 and 2022 International Energy Efficiency Scorecards. The slip in rankings is “despite a slight increase (3 points) in overall score,” ACEEE said in its report, “due to larger point gains among the top performers.” The four categories the group scores are national efforts, buildings, industry and transportation. Countries can earn up to 25 points in each section for a total of 100 points in subcategories such as fuel economy and emission standards or residential and commercial building codes. The U.S. in 2022 scored 16.5 points in national efforts, 17 in buildings, 12 in industry and 8.5 in transportation. In 2025 it scored 21 in national efforts, 17 in buildings, 13 in industry and 6 in transportation. Since 2022, “countries have made incremental progress in improving energy efficiency in buildings and industry,” ACEEE said. “However, even the top-performing countries have substantial room for improvement in advancing transportation efficiency.” “Nearly every country could work to significantly reduce travel from personal vehicles,” the report said. China improved its ranking from 9th in 2022 to sharing 5th place with Spain in this year’s report, thanks in part to its global leadership in public transit use and its national targets for reducing energy consumption and energy intensity. In 2025, the top four countries in order are France, Germany, the United Kingdom and Italy. France also ranked first in 2022. ACEEE cites France’s 2024 commitment of €800 million, or $932.9 billion, “to building retrofits … with additional investments in low-carbon transportation,” and its 2023 release of a “national water saving plan to cut water use by
WASHINGTON—U.S. Secretary of Energy Chris Wright today announced the Department of Energy’s (DOE) Loan Programs Office (LPO) closed a loan to support independent, American-made, and coal-powered fertilizer production. The $1.5 billion loan to Wabash Valley Resources, LLC, will help finance a coal and ammonia fertilizer facility in West Terre Haute, Indiana. The project will restart and repurpose a coal gasification plant idled since 2016 to produce 500,000 metric tons of anhydrous ammonia per year by using coal from a nearby Southern Indiana mine and petcoke as feedstock. “For too long, America has been dependent on foreign sources of fertilizer,” said U.S. Energy Secretary Chris Wright. “Under President Trump’s leadership, we are changing that by putting America first, relying on American coal, American workers, and American innovation to power our farms and feed our families.” By investing in a coal community, the Wabash project will bring the gasification plant back online to produce ammonia fertilizer – a vital resource for farmers across the Corn Belt, which currently relies on imports from Canada, the Caribbean, the Middle East, and Russia. The project will strengthen domestic supply chains, lower costs for farmers and consumers, and strengthen national food security by producing cost-competitive ammonia for the Eastern Corn Belt while creating hundreds of American jobs. The loan, which was carefully evaluated under the new LPO guidance directed by Secretary Wright, delivers on the Trump administration’s promise to responsibly steward taxpayer dollars and unleash American energy dominance. The Wabash financial close is the second closed loan under the Energy Dominance Financing (EDF) Program created by the Working Families Tax Cut, also known as the One Big Beautiful Bill Act. Today’s announcement highlights DOE’s commitment to achieving President Trump’s national security and energy dominance goals by securing domestic fertilizer supply for farmers in the Corn Belt
An Islamist insurgency that froze TotalEnergies SE’s $24.5-billion gas project in Mozambique four years ago is intensifying, just as the French oil major prepares to restart development. Militants affiliated with Islamic State have in recent months carried out raids across the northeastern Cabo Delgado province that hosts the Total project and another led by Exxon Mobil Corp. The number of attacks against civilians in the region has almost doubled in 2025 to the highest in years, according to the United Nations. TotalEnergies’ announcement last weekend that it’s restarting the project — seen as pivotal to transforming one of the world’s poorest nations — lifted Mozambique’s eurobonds 2.9% on Monday. At the same time, there are growing fears among local communities that the security situation near the project site at Afungi is deteriorating — with more than 90,000 people fleeing attacks since the last week of September. “Right now, people are living in fear,” Andrew Bogrand, a senior policy adviser at Oxfam America, said after a visit to the region that included a resettlement village adjacent to the project. “Folks in Quitunda, police officers, contractors — they don’t see how this project can work if people are concerned about security.” The militants have in recent weeks made incursions into both Palma — neighboring the LNG site — and nearby Mocimboa da Praia, where they filmed themselves preaching in a local mosque. That was symbolic: The port town is where the extremist insurgency began in 2017 as a ragtag army of local youth in one of the poorest parts of Mozambique. They later occupied the town for about a year. Total has estimated charges of $4.5 billion during the halt to construction, adding to the original $20 billion project cost — revisions that require government approval before work resumes. The company paid
Indian Oil Corp. plans to start a joint venture with commodities trader Vitol Inc. in Singapore to trade oil and fuel products early next year, a person familiar with the development said. The agreement will include a clause requiring Vitol’s exit after five to seven years, said the person, who asked not to be named as talks aren’t public. The New Delhi-based refiner had held talks with several other companies, including BP Plc and TotalEnergies SE, but finally decided to move ahead with Vitol, the person said. Global trading giants have lost some hold over India’s market as the country turned to cheaper Russian barrels for its spot purchases. With India expected to lead global oil demand growth, they are now trying to regain ground. Teaming up with Indian Oil could give Vitol a stronger presence in the country as it cuts back on Russian oil due to sanctions from the West. Indian Oil and Vitol didn’t respond to messages seeking comment. Indian Oil will capitalize on the strength of the global trader which has access to real-time market intelligence, in addition to its wider reach and established risk management systems. That may help the Indian company source crude at lower prices. The refiner, which meets nearly 90% of its oil demand via imports, is likely to see a surge in consumption as its crude processing capacity is expected to expand by 346,000 barrels a day to 1.76 million barrels next year. WHAT DO YOU THINK? Generated by readers, the comments included herein do not reflect the views and opinions of Rigzone. All comments are subject to editorial review. Off-topic, inappropriate or insulting comments will be removed.
Dive Brief: The Federal Energy Regulatory Commission on Monday rejected a large load tariff proposed by Tri-State Generation and Transmission Association, saying it intruded on retail rate regulation, which falls under state jurisdiction and is outside of FERC’s authority. “We find that Tri-State has not provided a sufficient basis for the Commission to find that its proposal does not regulate the terms and conditions of a [High Impact Load] Customer’s retail service in ways that are beyond the Commission’s authority,” FERC said. With FERC beginning to consider the U.S. Department of Energy’s request that the agency develop rules for interconnecting large loads to the transmission system, the Tri-State decision provides insights into the commission’s thinking about its jurisdiction over the issue, according to Steven Shparber, a member at the law firm of Mintz, Levin, Cohn, Ferris, Glovsky and Popeo. “FERC’s decision relied on longstanding U.S. Supreme Court precedent noting that FERC may not regulate retail sales, which are exclusively within the states’ jurisdiction,” he said in an email. Dive Insight: Tri-State’s Aug. 28 proposal came as utilities like AEP Ohio and Dominion Energy Virginia have adopted tariffs that set out terms and conditions for interconnecting data centers and other large loads. In its proposal, Tri-State — a wholesale cooperative based in Westminster, Colorado, with 40 utility members in Colorado, Nebraska, New Mexico and Wyoming — said there is growing interest in building data centers in the Mountain West region. It has received 10 requests from member utilities to add loads ranging from 45 MW to 650 MW that could each grow to 300 MW to 1,000 MW, according to Tri-State, which has a 2,500 MW peak load. Tri-State said its proposed High Impact Load, or HIL, tariff and High Impact Load Agreement would protect its utility members from a range of risks
The enormous amount of energy required to power a modern data center has prompted major tech companies to sign major partnership agreements with power companies. Most recently, Google signed an agreement with Next Era Energy to restart the Duane Arnold Energy Center in Iowa. The nuclear power plant in question was shut down in 2020 and it is expected to take four years to make it operational again, CNBC reports.
The 7280R4-32PE features 25.6 Tbps switching capacity and supports 32x 800 GbE ports with Octal Small Form-Factor Pluggable (OSFP) or Quad Small Form-Factor Pluggable – Double Density (QSFP-DD) optical uplinks. It’s targeted at customers that need to support AI/ML workloads and routing-intensive edge use cases, Arista stated. It supports 25% lower power per Gbps compared to the prior generation, according to Arista. A second version, the 7280R4-64QC-10PE, is aimed at dense, deep buffer-requiring workloads in data centers with 100G/800G requirements. The box supports 64x 100 GbE and 10x 800 GbE OSFP in addition to 4x 1/10/25 GbE for management or additional low-speed interfaces, Arista started. The box promises 20% lower power requirement per Gbps over the prior generation of the box, Arista stated. At the high end, the new 7800R4 is the vendor’s latest flagship networking box capable of supporting 36 ports of 800GbE OSFP and QSFP-DD line cards in 4, 8, 12, and 16-slot chassis configurations. The box offers a high radix capacity – meaning it can be fully loaded with line card and support 576 physical 800 Gigabit Ethernet ports or 1,152 400GbE ports, Arista stated. In addition, the 7800R supports a new 3.2 TbpsEthernet line card called HyperPort that supports 4 800G channels to tie together widely dispersed data centers via a technique Arista calls “scale across.” It’s designed to scale across buildings in the same metropolitan region or across sites in different cities or countries. This routed Data Center Interconnect technology that can extend AI clusters over Metro or long-haul WAN links, according to Arista. “Building on the flexible Extensible Operating System (EOS) software foundation [which runs across all Arista networking gear] and deep buffering, HyperPort delivers up to 44% faster job completion time (JCT) for high-bandwidth AI flows via a single high-speed port, compared to
The new box extends Cisco Nexus 9000 Series portfolio of high-density 800G aggregation switches for the data center fabric, Cisco stated. The Nexus 9000 data center switches are a core component of the vendor’s enterprise AI offerings. They support congestion-management and flow-control algorithms and deliver the right latency and telemetry to meet the design requirements of AI/ML fabrics, Cisco stated. With the Cisco N9100 Series, Cisco now supports Nvidia Cloud Partner (NCP)-compliant reference architecture. “This development is particularly significant for neocloud and sovereign cloud customers building data centers with capacities ranging from thousands to potentially hundreds of thousands of GPUs, as it allows them to diversify their supply chains effectively,” wrote Will Eatherton, senior vice president of Cisco networking engineering, in a blog post about the news. An add-on license lets customers extend the NCP reference architecture to define how customers can mix and mingle Nvidia Spectrum-X adaptive routing capability with Cisco Nexus 9300 Series switches and Nvidia Spectrum-X Ethernet SuperNICs. “The combination of low latency and congestion-aware, per-packet load balancing on Cisco 9300 switches, along with out-of-order packet handling and end-to-end congestion management on Nvidia SuperNICs, significantly enhances network performance. These improvements are essential for AI networks, optimizing critical metrics such as job completion time,” Eatherton wrote. In addition to neoclouds and sovereign buildouts, enterprise customers are a target, according to Futuriom’s Raynovich.
Organizations must ensure the infrastructure is AI ready Infrastructure is another area where Cisco found a major difference. Pacesetters are designing their networks for future demands. Seventy-one percent say their networks can scale instantly for new AI projects. Roughly three-quarters of pacesetters are investing in new data center capacity over the next year. Currently, about two-thirds say their infrastructure can accommodate AI workloads. Most pacesetters (93%) also have data systems that are fully prepared for AI, compared with 34% of other companies. About 76% have fully centralized their in-house data, while only 19% of other companies have done the same. Eighty-four percent report strong governance readiness, while 95% have mature processes to measure the impact of AI. If ever there was a technological shift that requires the right infrastructure, it’s AI. AI generates a significant amount of data, needs large amounts of processes and low latency, high-capacity networks. Historically, businesses could operate with networks that operated on the premise of “best effort,” but that’s no longer the case. From the data center to campus to branch offices, in most companies, the network will require a refresh. Scaling AI requires the right processes When it comes to being disciplined, 62% of pacesetters have an established process for generating, piloting, and scaling AI use cases. Only 13% of other organizations (non-pacesetters) have reached this level of maturity. Most pacesetters say their AI models achieve at least 75% accuracy. Almost half also expect a 50% to 100% return on investment (ROI) within a year, far above the average. Cisco notes that over the past six months, pressure has been building for companies to show tangible ROI. Executives and IT leaders are pushing for results, and so are competitors. By contrast, most other companies are in early stages of readiness. Although 83% plan to
From a strategy perspective, there is a longer term enterprise play here, noted Moor’s Kimball; Humain is Qualcomm’s first customer, and a cloud service provider (CSP) or hyperscaler will likely be customer number two. However, at some point, these rack-scale systems will find their way into the enterprise. “If I were the AI200 product marketing lead, I would be thinking about how I demonstrate this as a viable platform for those enterprise workloads that will be getting ‘agentified’ over the next several years,” said Kimball. It seems a natural step, as Qualcomm saw success with its AI100 accelerator, a strong inference chip, he noted. Right now, Nvidia and AMD dominate the training market, with CUDA and ROCm enjoying a “stickiness” with customers. “If I am a semiconductor giant like Qualcomm that is so good at understanding the performance-power balance, this inference market makes perfect sense to really lean in on,” said Kimball. He also pointed to the company’s plans to re-enter the datacenter CPU space with its Oryon CPU, which is featured in Snapdragon and loosely based on technology it acquired with its $1.4 billion Nuvia acquisition. Ultimately, Qualcomm’s move demonstrates how wide open the inference market is, said Kimball. The company, he noted, has been very good at choosing target markets and has seen success when entering those markets. “That the company would decide to go more ‘in’ on the inference market makes sense,” said Kimball. He added that, from an ROI perspective, inferencing will “dwarf” training in terms of volume and dollars.
However, that’s only one part of the problem. Meeting the power demands of AI data centers will require the energy sector to make large investments. Then there’s data center demand for microprocessors, rare earth elements, and other valuable metals such as copper, which could, in a bust, make data centers the most expensively-assembled unwanted assets in history. “Financial stability consequences of an AI-related asset price fall could arise through multiple channels. If forecasted debt-financed AI infrastructure growth materializes, the potential financial stability consequences of such an event are likely to grow,” warned the BoE blog post. “For companies who depend on the continued demand for massive computational capacity to train and run inference on AI models, an algorithmic breakthrough or other event which challenges that paradigm could cause a significant re-evaluation of asset prices,” it continued. According to Matt Hasan, CEO of AI consultancy aiRESULTS, the underlying problem is the speed with which AI has emerged. “What we’re witnessing isn’t just an incremental expansion, it’s a rush to construct power-hungry, mega-scale data centers,” he told Network World. The dot.com reversal might be the wrong comparison; it dented the NASDAQ and hurt tech investment, but the damage to organizations investing in e-commerce was relatively limited. AI, by contrast, might have wider effects for large enterprises because so many have pinned their business prospects on its potential. “Your reliance on these large providers means you are indirectly exposed to the stability of their debt. If a correction occurs, the fallout can impact the services you rely on,” said Hasan.
And Microsoft isn’t the only one that is ramping up its investments into AI-enabled data centers. Rival cloud service providers are all investing in either upgrading or opening new data centers to capture a larger chunk of business from developers and users of large language models (LLMs). In a report published in October 2024, Bloomberg Intelligence estimated that demand for generative AI would push Microsoft, AWS, Google, Oracle, Meta, and Apple would between them devote $200 billion to capex in 2025, up from $110 billion in 2023. Microsoft is one of the biggest spenders, followed closely by Google and AWS, Bloomberg Intelligence said. Its estimate of Microsoft’s capital spending on AI, at $62.4 billion for calendar 2025, is lower than Smith’s claim that the company will invest $80 billion in the fiscal year to June 30, 2025. Both figures, though, are way higher than Microsoft’s 2020 capital expenditure of “just” $17.6 billion. The majority of the increased spending is tied to cloud services and the expansion of AI infrastructure needed to provide compute capacity for OpenAI workloads. Separately, last October Amazon CEO Andy Jassy said his company planned total capex spend of $75 billion in 2024 and even more in 2025, with much of it going to AWS, its cloud computing division.
Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Self-driving tractors might be the path to self-driving cars. John Deere has revealed a new line of autonomous machines and tech across agriculture, construction and commercial landscaping. The Moline, Illinois-based John Deere has been in business for 187 years, yet it’s been a regular as a non-tech company showing off technology at the big tech trade show in Las Vegas and is back at CES 2025 with more autonomous tractors and other vehicles. This is not something we usually cover, but John Deere has a lot of data that is interesting in the big picture of tech. The message from the company is that there aren’t enough skilled farm laborers to do the work that its customers need. It’s been a challenge for most of the last two decades, said Jahmy Hindman, CTO at John Deere, in a briefing. Much of the tech will come this fall and after that. He noted that the average farmer in the U.S. is over 58 and works 12 to 18 hours a day to grow food for us. And he said the American Farm Bureau Federation estimates there are roughly 2.4 million farm jobs that need to be filled annually; and the agricultural work force continues to shrink. (This is my hint to the anti-immigration crowd). John Deere’s autonomous 9RX Tractor. Farmers can oversee it using an app. While each of these industries experiences their own set of challenges, a commonality across all is skilled labor availability. In construction, about 80% percent of contractors struggle to find skilled labor. And in commercial landscaping, 86% of landscaping business owners can’t find labor to fill open positions, he said. “They have to figure out how to do
Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More 2025 is poised to be a pivotal year for enterprise AI. The past year has seen rapid innovation, and this year will see the same. This has made it more critical than ever to revisit your AI strategy to stay competitive and create value for your customers. From scaling AI agents to optimizing costs, here are the five critical areas enterprises should prioritize for their AI strategy this year. 1. Agents: the next generation of automation AI agents are no longer theoretical. In 2025, they’re indispensable tools for enterprises looking to streamline operations and enhance customer interactions. Unlike traditional software, agents powered by large language models (LLMs) can make nuanced decisions, navigate complex multi-step tasks, and integrate seamlessly with tools and APIs. At the start of 2024, agents were not ready for prime time, making frustrating mistakes like hallucinating URLs. They started getting better as frontier large language models themselves improved. “Let me put it this way,” said Sam Witteveen, cofounder of Red Dragon, a company that develops agents for companies, and that recently reviewed the 48 agents it built last year. “Interestingly, the ones that we built at the start of the year, a lot of those worked way better at the end of the year just because the models got better.” Witteveen shared this in the video podcast we filmed to discuss these five big trends in detail. Models are getting better and hallucinating less, and they’re also being trained to do agentic tasks. Another feature that the model providers are researching is a way to use the LLM as a judge, and as models get cheaper (something we’ll cover below), companies can use three or more models to
Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More OpenAI has taken a more aggressive approach to red teaming than its AI competitors, demonstrating its security teams’ advanced capabilities in two areas: multi-step reinforcement and external red teaming. OpenAI recently released two papers that set a new competitive standard for improving the quality, reliability and safety of AI models in these two techniques and more. The first paper, “OpenAI’s Approach to External Red Teaming for AI Models and Systems,” reports that specialized teams outside the company have proven effective in uncovering vulnerabilities that might otherwise have made it into a released model because in-house testing techniques may have missed them. In the second paper, “Diverse and Effective Red Teaming with Auto-Generated Rewards and Multi-Step Reinforcement Learning,” OpenAI introduces an automated framework that relies on iterative reinforcement learning to generate a broad spectrum of novel, wide-ranging attacks. Going all-in on red teaming pays practical, competitive dividends It’s encouraging to see competitive intensity in red teaming growing among AI companies. When Anthropic released its AI red team guidelines in June of last year, it joined AI providers including Google, Microsoft, Nvidia, OpenAI, and even the U.S.’s National Institute of Standards and Technology (NIST), which all had released red teaming frameworks. Investing heavily in red teaming yields tangible benefits for security leaders in any organization. OpenAI’s paper on external red teaming provides a detailed analysis of how the company strives to create specialized external teams that include cybersecurity and subject matter experts. The goal is to see if knowledgeable external teams can defeat models’ security perimeters and find gaps in their security, biases and controls that prompt-based testing couldn’t find. What makes OpenAI’s recent papers noteworthy is how well they define using human-in-the-middle
This ebook is available only for subscribers.
In this exclusive subscirber-only ebook you’ll learn how the emissions from individual AI text, image, and video
In partnership withDatabricks Four years is a lifetime when it comes to artificial intelligence. Since the first edition of this study was published in 2021,
Mathematics is the foundational language of the universe, providing the tools to describe everything from the laws of physics to the intricacies of biology and
This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. DeepSeek
